lsb

23 ms·

151.
by lsb 6y ago
"What did you do yesterday", from the article, is a great open-ended question that tells you a lot about someone. I've always enjoyed a diff-question, like, what do you wish people knew about xyz, or what surprised you about
152.
by lsb 6y ago
A lot of the data comes from NYC open data, and the results are gorgeous (especially the hydrology) See previously: https://news.ycombinator.com/item?id=18837481
153.
Mapping and Visualization(scottreinhard.com)
1 points by lsb 6y ago|1 comments
154.
Selfie to Waifu(waifu.lofiu.com)
1 points by lsb 6y ago|0 comments
155.
by lsb 6y ago
This is 10 years old! The GPU the benchmarks run on has 480 cores, and modern GPUs have >5000, and the available memory has gone from 1.5GB to 32GB. I'd love to see these numbers run with a modern GPU.
156.
by lsb 6y ago
Reminds me of when someone saw that the NSA's public key ( http://www.cypherspace.org/adam/hacks/lotus-nsa-key.html ) to be included in Lotus Notes had an organizational name of "MiniTruth", and a com
157.
by lsb 6y ago
I'm pretty sure that's a great new botanical OnlyFans
158.
by lsb 7y ago
This looks super interesting! 1) How does this compare to BlazingDB/MapD/etc, especially in queries on <some benchmark> per $ per hour on AWS? 2) I don't see a way to give you money. What ensures that QuestDB will be ar
159.
by lsb 7y ago
I recently switched my phone to route all internet traffic through Tor and the DDG browser on Android, but the hCaptcha challenges are even more exasperating than the ReCaptchas, and I've found my bounce rate close to 100%. Is the abun
160.
by lsb 7y ago
Acceleration. The rate of change of the rate of change of deaths. The word exponential only occurs once in the paper, and the rate of change of an exponential is an exponential, so take as many derivatives as you want and it's still go
161.
by lsb 7y ago
That 0.01 millirem is roughly 1 Banana Equivalent Dose, the radiation you get from eating a banana. https://en.wikipedia.org/wiki/Banana_equivalent_dose
162.
by lsb 7y ago
"a worldwide planting programme could remove two-thirds of all the emissions from human activities that remain in the atmosphere today" https://www.theguardian.com/environment/2019/jul/04/planti
163.
by lsb 7y ago
When I tried loading the page, it timed out, whereas now I'm getting an immediate 451
164.
by lsb 7y ago
Oh no! I'm seeing it down. https://downforeveryoneorjustme.com/whereisscihub.now.sh Sending HugOps to everyone at Zeit trying to keep 99.99% uptime for now.sh at a time like this
165.
by lsb 7y ago
Does that mean that answer grading would become like comparing summaries of a given text?
166.
by lsb 7y ago
What do you like about the latent space that you are sampling from? :)
167.
by lsb 7y ago
You can get a box with 4TB of ram on EC2 for $4/hr spot, so copy your data into /dev/shm and go hog wild. For lots of databases, most of their time is spent locking and copying data around, so depending on your workload you m
168.
by lsb 7y ago
I think mauve has the most RAM
169.
by lsb 7y ago
The model has 18x512 parameters, and all 9000 parameters are 32-bit floats. Even assuming only 16 bits of randomness for each parameter (to keep them small enough, so you don't get too wild, because the center of 0 is a pretty homogeni
170.
by lsb 7y ago
StyleGAN (v2, which is this, and v1 before it) is spectacularly high quality. Even more interestingly, you can interpolate between faces very smoothly. (For instance, I was able to give some face off the street a drag makeover, by interpola
171.
by lsb 7y ago
Author here. Happy to answer any questions :)
172.
Show HN: Using Pandas/WebGL/SQLite to map a million new sister cities(leebutterman.com)
12 points by lsb 7y ago|1 comments
173.
by lsb 7y ago
Does the Numba JIT do this? This definitely seems like a compiler optimization that I want to mechanically run on my code
174.
by lsb 7y ago
That's why you can `EXPLAIN` your query, and maybe even add indices, to speed things up! Plus in the last few years SQLite has gotten a lot more powerful, and has added lots of JSON support and has improved index usage. I'd avoid
175.
by lsb 7y ago
> The culprit was insufficient testing The culprit was not insufficient testing! The culprit was insufficient testing, plus lacks of restorable backups, on multiple levels. Last level of backup: there's 5 million records, use a
176.
by lsb 7y ago
CODE_OF_CONDUCT.md Examples of unacceptable behavior by participants include: Publishing others' private information, such as a physical or electronic address, without explicit permission How is this tool not a violation of its o
177.
by lsb 7y ago
That's a great phrase from https://blog.janestreet.com/effective-ml-revisited/
178.
by lsb 7y ago
Yes! People have trained large birds to do this as well! https://duckduckgo.com/?q=hawk+catching+a+drone&ia=videos&ia...
179.
by lsb 7y ago
It's the subtlety, because vision algorithms don't necessarily look at the world like we do. Here, some folks perturbed an image of a cat to scan as guacamole, and have 3d printed a model of a turtle that scans as a rifle: https:
180.
by lsb 7y ago
I've always enjoyed the multiple valid parses of "Time flies like an arrow". I can't wait for AI to generate more Escher sentences like "More people have been to Russia than I have" ( https://en.m.wi

More ›